I want to start off by linking to my post mortem for our previous game, Fall of Atlantis. If you are an independent developer like ourselves or are looking to get into the business there are still tons of wisdom in there.
So it has been almost 6 months and our second game is finally out on the App Store. It has been a long ride and to be honest it hasn't been the success we expected it to be. The media gave us generally favorable reviews but the public hasn't been as kind with sales starting to dwindle after the first day. However as always there are great lessons to be learned... in fact it is always the guy who didn't quite make it who becomes the teacher of the next generation.
Hopefully we will make it anyway even though I'm writing this. Haha.
What went right
1. Throw it in the bin and do it again
After finishing the initial version of our game we received some playtesting that basically said: "hey, this game is utter crap". So with only 3 weeks to go we scrapped the whole game and pulled ourselves together to make a whole new game. We had about 2 months of development on the game total.
Obviously we tried to reuse as much as possible but to illustrate how much we actually scrapped.. well.. the initial game was more in the line of a racing game while the end product is more of a Metroidvania with exploration, powerups and all that jazz.
This might've sounded like the recipe for disaster but surprisingly it worked out and we had a lot of interest and even praise among iPhone gaming media. The game obviously has quite a few shortcomings regarding the gameplay but I think it was well done considering the time we had to make the game on.
2. Congratulations! You leveled up in Marketing
In our last post mortem we mentioned that we had a huge issue with marketing; in fact it was non-existant! As it could not have possibly been worse it did kind of feel good to actually start working on marketing this time around. I'm not sure if the marketing went well enough to be placed here but I'm too proud to not.
As this was our first real attempt to market ourselves we had a lot of stuff to do. The most important thing you can do when marketing is research! Where do your customers go when they want to buy a game? What are they interested in? How much do they want to pay for the game and how do you make them pay more? All these questions are obvious but it is really really hard to find out the answers.
One thing we did was to actually listen to our testers. What was it that they reacted on? Was this something we could push when making our marketing? One thing we were unsure of was the quality of the graphics but everyone said that the game looked great so we opted to push on that... and surely everyone picked it up. When we got reviews the reviewers were keen to point out that we had lovely graphics. To this date I'm still not sure if we do but hey, I'm not the one buying the game!
Oh right, this Gamasutra feature was very helpful too. It helped us identify who had actually written about us so we could contact them as well as mention them on our blog.
http://www.youtube.com/watch?v=AIHHS46NBzs - Marketing in the work. This video was quite fun to make and closely copied leading iPhone game trailers.
3. Using references
After scrapping the game and deciding to make it more of a Metroidvania game we figured it would actually bewise to look at these type of games. I was the level designer so I promptly sat down and played Castlevania: Symphony of the Night as I had never played these type of games before. It took me a while and I had to play it with a map in front of me in order to study it properly within my time limits. In the end I think the level design gets a passing grade and it was quite influenced by Castlevania.
The other thing we looked at was how these games do their world maps and we realized that they are very simple. Before we looked at them we had planned to make a map that was based around aesthetics instead of functionality and in the end these functional maps are a lot better. I mean... these games are about exploring the map not admiring it's looks.
Metroid map stolen! (actually it was stolen from Shadow Complex)
4. Dynamic Music
This was a slight experiment for us. We knew that the guy who made our music was already up to par with our competitor's music so we threw him something else we wanted to try: dynamic music. Now it may sound like an easy task but it was actually riddled with problems as the iPhone is very bad at playing multiple compressed sound files without interruption. There was a lot of stuttering going on.
In the end we had to opt for uncompressed music and that limited how much music we could have. However the end result is pleasing and people have praised the music a lot.
5. Art direction also levels up
Here's another thing that we had problems with in our last game. As mentioned already people have actually said that the game looks good. This was mostly due to us choosing a theme very early on which was: fairytale forest with a Ghibli feeling to it. I'm not going to claim it actually looked like that in the end but a unifying theme is important!What went wrong
1. Pre-production is still serious business
We never seem to learn. Pre-production is probably the most important stage in game development. It is all about eliminating risks.. and we already knew that from our last post mortem. So why oh why do I have to mention this again? Shouldn't we have learnt our lesson already? To be honest it may be because the perfect pre-production does not exist.. it is just not possible cause the reason you have a pre-production is because things can't go perfectly as planned ever.
Our biggest mistake was to think that prototyping is only limited to gameplay mechanics. That was basically the only thing we spent time on to prototype... and even then we didn't bother with ironing things out before moving on.
Pre-production takes time so you need to choose a project with the right amount of risks that you are going to prototype. You can't be making your dream game right from the get go cause you just won't have time enough to prototype everything at once. You would most likely be spending five years in pre-production only...
2. Not using references
This is related to the pre-production issue. The only surefire way of avoiding risks is doing something that has already been done... cause the only risk is that you are making a game that is boring cause it has been done to death. However most of of the time this won't be a problem at all. I mean just think about designing how an explosion should look... why waste time on that when it has already been done so many times. Sure, it will look a little copied but time is precious and with a small team you will not have time to poke in every little detail. Always look for opportunities to steal things from others. No one is going to notice anyway.
Our biggest fault was that the first version of our game didn't really have a predecessor for us to look at... or.. it probably did but then it must've been as bad as our version and well.. it wouldn't have been that worthwhile to dig up and look at... timewise that is (I mean it is always good to learn from other's mistake).
3. Closed testing isn't perfect
Something that actually went well that I didn't mention was our playtesting. We managed to squash quite a few bugs with that. However we kind of went blind and thought that with testers we would know exactly how good the game was and how the public would react. BEEP! WRONG! Closed testing is good for finding bugs and stuff but knowing how people will react when playing the game for the first time however is another question.
In our case people initially had issues with our control scheme but after a few tweaks and tests it started to disappear so we were content... but obviously it wasn't just that our controls got better... it was that our testers were getting used to them! So when the game was released we had a lot of people cry to us that the controls were unintuitive or that we were lacking proper optional control schemes.
4. Standards are there for a reason
As mentioned above our control scheme received quite a few complains.. or rather the lack of optional control schemes was the problem. So after a few searches later we realized that most games with esoteric control schemes also have a more traditional way of letting the player control the game.
Well, we missed that but we have added it to our list of things to research. Always look up what everyone else does and try to realize why they do it. You can never make a game that is above others if you do not know what others do that is good.
Oh, just to be clear.. even if the standard makes your game slightly worse it may be very well worth to implement a more standard version of your feature cause well.. it may be what people actually want. I mean... no one wants to spend 10min to learn to move around the level.
5. Wow! It loks good!.... but what is it?
Another thing that compounded to the control confusion was the fact that at the end the graphics looked good but... no one could tell what it actually was. People actually thought that the main character was a beaver swimming underwater fighting fish monsters.. and well.. I can't blame them.
This wouldn't actually have been a problem cause hey... if it looks good it will still look good and sell your game. However since people thought they were swimming they didn't understand our hanggliding mechanics of gaining speed and moving around the level.
If you've watched the trailer above you will probably see what I mean... it was supposed to be a fairytale forest though!
Is it a beaver? Squirrel? Maybe even a sea lion?!
So yeah, we are quite happy with the results especially considering that we scrapped the whole gameplay with only 3 weeks to go. Hopefully I'll be back making more post mortems...
If you have any comments, questions or concerns please post them. We are constantly trying to improve and we feel that we won't be able to do that without proper feedback.